Imaging mass cytometry (IMC) enables the in situ analysis of in-depth-phenotyped cells in their native microenvironment within the preserved architecture of a single tissue section. To date, it permits the simultaneous analysis of up to 50 different protein- markers targeted by metal-conjugated antibodies. The application of IMC in the field of cancer research may notably help 1) to define biomarkers of prognostic and theragnostic significance for current and future treatments against well-established and novel therapeutic targets and 2) to improve our understanding of cancer progression and its resistance mechanisms to immune system and how to overcome them. In the present article, we not only provide a literature review on the use of the IMC in cancer-dedicated studies but we also present the IMC method and discuss its advantages and limitations among methods dedicated to deciphering the complexity of cancer tissue.

Persistent inflammation can promote the development of tertiary lymphoid structures (TLS) within tissues resembling secondary lymphoid organs (SLO) such as lymph nodes (LN). The composition of TLS across different organs and diseases could be of pathophysiological and medical interest. In this work, we compared TLS to SLO in cancers of the digestive tract and in inflammatory bowel diseases. Colorectal and gastric tissues with different inflammatory diseases and cancers from the department of pathology of CHU Brest were analyzed based on 39 markers using imaging mass cytometry (IMC). Unsupervised and supervised clustering analyses of IMC images were used to compare SLO and TLS. Unsupervised analyses tended to group TLS per patient but not per disease. Supervised analyses of IMC images revealed that LN had a more organized structure than TLS and non-encapsulated SLO Peyer’s patches. TLS followed a maturation spectrum with close correlations between germinal center (GC) markers’ evolution. The correlations between organizational and functional markers made relevant the previously proposed TLS division into three stages: lymphoid-aggregates (LA) (CD20+CD21-CD23-) had neither organization nor GC functionality, non-GC TLS (CD20+CD21+CD23-) were organized but lacked GC’s functionality and GC-like TLS (CD20+CD21+CD23+) had GC’s organization and functionality. This architectural and functional maturation grading of TLS pointed to differences across diseases. TLS architectural and functional maturation grading is accessible with few markers allowing future diagnostic, prognostic, and predictive studies on the value of TLS grading, quantification and location within pathological tissues in cancers and inflammatory diseases.

The recent emergence of imaging mass cytometry technology has led to the generation of an increasing amount of high-dimensional data and, with it, the need for suitable performant bioinformatics tools dedicated to specific multiparametric studies. The first and most important step in treating the acquired images is the ability to perform highly efficient cell segmentation for subsequent analyses. In this context, we developed YOUPI (Your Powerful and Intelligent tool) software. It combines advanced segmentation techniques based on deep learning algorithms with a friendly graphical user interface for non-bioinformatics users. In this article, we present the segmentation algorithm developed for YOUPI. We have set a benchmark with mathematics-based segmentation approaches to estimate its robustness in segmenting different tissue biopsies.

Objective: Persistent inflammation can promote the development of tertiary lymphoid structures (TLS) within tissues resembling the secondary lymphoid organs (SLO) as lymph nodes (LN). The composition of the TLS across different organs and diseases could be of pathophysiological and medical interest. In this work, we compared TLS to SLO and between cancer and inflammatory diseases of the digestive tract. Design: Colorectal and gastric tissues with different inflammatory diseases and cancers from the department of pathology of CHU Brest were analyzed based on 39 markers using imaging mass cytometry (IMC). Unsupervised and supervised clustering analyses of IMC images were used to compare SLO and TLS. Results: Unsupervised analyses tended to group TLS per patient but not per disease. Supervised analyses of IMC images revealed that LN had a more organized structure than TLS and non-encapsulated SLO Peyer's patches. TLS followed a maturation spectrum with close correlations between germinal cell (GC) markers' evolution. The correlations between organizational and functional markers made relevant the previously proposed TLS division into three stages: lymphoid-aggregates (LA) (CD20+CD21-CD23-) had neither organization nor GC functionality, non-GC TLS (CD20+CD21+CD23-) were organized but lacked GC's functionality and GC-like TLS (CD20+CD21+CD23+) had GC's organization and functionality. This architectural and functional maturation grading of TLS pointed to differences across diseases. Conclusion: TLS architectural and functional maturation grading is accessible with few markers allowing future diagnostic, prognostic, and predictive studies on the value of TLS grading, quantification and location within pathological tissues in cancers and inflammatory diseases.